论文部分内容阅读
针对目前多数故障检测法对缓变的软故障检测不灵敏,延迟性较大,无法判断各种故障类别等问题,在分析利用小波奇异性进行故障检测原理的基础上,提出了一种基于模极大值原理的导航传感器故障检测方法。该算法利用传感器的观测量来诊断传感器是否正常工作,文中给出了利用小波奇异性进行故障检测的原理描述及利用该算法进行故障检测算法流程。该算法不仅对导航传感器软故障具有较高的检测灵敏度,而且可以通过故障点处信号的Lipschitz指数来判断故障类别,为传感器故障隔离和修复提供有效的信息,因此可以更好的保障多传感器容错组合导航系统的可靠性。计算机仿真表明所提方法是有效的,具有较高的实用性。
In view of the fact that most fault detection methods are insensitive to slow-grained soft fault detection and large in delay, it is impossible to judge various fault types and so on. On the basis of analyzing the principle of fault detection using wavelet singularity, Maximum Sensor Fault Detection Based on Maximum Principle. The algorithm uses sensor observations to diagnose whether the sensor is working properly. In this paper, the principle of fault detection using wavelet singularity is given and the flow of algorithm for fault detection is given. The algorithm not only has high detection sensitivity for soft sensor navigation faults, but also can determine the fault type through the Lipschitz index of the signal at the fault point, which can provide effective information for sensor fault isolation and repair, so it can better protect the multi-sensor fault tolerance Integrated navigation system reliability. Computer simulation shows that the proposed method is effective and has high practicability.